Personalization for Learning Algebraic Symbolization in an AI-Based Environment

• In Event: Algebra, Proportional Reasoning, and Working Memory

Abstract

The context of computer-based learning offers unique affordances and challenges
for students who come from diverse backgrounds and interests. This necessitates
expanding an understanding about these students’ perceptions and learning
experiences in these technology-enhanced contexts. We analyze observations
about algebra-learning behaviors from interviews, think-aloud, and observation
sessions with students using an AI Tutoring System. Findings reveal that low
interest in learning symbolization influenced students' judgment of and interest in
using the AI tutor. Implications include an improved design for the AI tutor to
provide varying levels of motivational interventions to support students’
growth-mindset and self-efficacy in symbolization. Additionally, enhanced
scaffolding through personalized grounded feedback and utility-value intervention
designed specifically for learning to create symbolic models of problem scenarios

Authors

• Pallavi Chhabra, Washington University in St. Louis

• Danielle R. Thomas, Carnegie Mellon University

• Kenneth R. Koedinger, Carnegie Mellon University

• Vincent Aleven, Carnegie Mellon University